Scheduling Remote Access to Scientific Instruments in Cyberinfrastructure for Education and Research

While a grid represents a computing infrastructure for cross domain sharing of computational resources, the cyberinfrastructure, proposed by the US NSF Blue - Ribbon advisory panel, is expected to revolutionizing science and engineering by including more computer integrated resources, e.g. telescopes and observatories. As a part of the China national cyberinfrastructure for education and research, resource sharing of expensive scientific instruments is discussed in this work. A layered model of instrument pools is introduced and the process from submitting a job to instrument pools to obtaining results is analyzed. Fuzzy random scheduling algorithms are proposed in instrument pools when a job is submitted to one of instruments within a pool. The randomness lies in the probability which instrument could be chosen for an experiment and the fuzziness origins from vagueness of users' feedback opinions on experimental results. Users' feedback information is utilized to improve overall quality of service (QoS) of an instrument cyberinfrastructure. Several algorithms are provided to increase utilization of instruments providing higher QoS and decrease utilization of those with poor QoS. This is demonstrated in details using quantitative simulation results included in this paper.